analytic
harmonic
functions
Overview
Does the notion of a derivative of a complex function make sense? If so, how
should it be defined and what does it represent? These and similar questions
are the focus of this chapter. As you might guess, complex derivatives have a
meaningful definition, and many of the standard derivative theorems from calculus (such as the product rule and chain rule) carryover. There are also some
interesting applications.
But not everything is symmetric. You will learn in
this chapter that the mean value theorem for derivatives does not extend to
complex functions. In later chapters you will see that differentiable complex
functions are, in some sense. much more "differentiable" than differentiable
real functions.
3.1
DIFFERENTIABLE
FUNCTIONS
AND ANALYTIC
Using our imagination, we take our lead from elementary calculus and define the
derivative of f at zo, written J' (zo), by
f / (zo)
lim
f (z)  f (zo) ,
Z 
Z>ZQ
Zo
provided the limit exists. If it does, we say that the function f is differentiable
at zoo If we write 6.z = z  zo, then we can express Equation (31) in the form
I (
Zo 
l'
1m
Ll.z>o
uZ
dw
1 (zo) = dz =
I' (zo) =
6.w
lim ~.
Z3 
L>Z>O DoZ
lim z>zo
1(z)  1(zo),
z5
z  Zo
(zzo)
= hm z>zo
(z2+zoz+z5)
z  Zo
= 3z5
3.2
1~
1~
1(z)
= z =
x  iy is nowhere
Solution
We choose two approaches to the point Zo = xo + iyo and compute
limits of the difference quotients. First, we approach Zo = xo + iyo along a line
parallel to the xaxis by forcing z to be of the form z = x + iyo.
lim
Z>Zo
_1_(z_)__I_(_zo_)
Z  Zo
lim
(x+iyo)>(xo+iyo)
.
11m
(x+iyo)>(xo+iyo)
lim
(x+iYo)>(xo+iYoJ


x  Xo
f(z)f(zo)
11m "'
Z  Zo
Z~ZO
l'
1m
(xo+iy)~(xo+iyo)
.
11m
(xo+iy)~(xo+iyo)
lim
(xo+iy)~(xo+iyo)
f(xo+iy)f(xo+iyo)
(xo + iy)  (xo + iyo)
(xo  iy)  (xo  iyo)
(xo  xo) + i (y  Yo)

i (y  Yo)
i (y  Yo)
= 1.
The limits along the two paths are different, so there is no possible value for
the right side of Equation (31). Therefore, f (z) = z is not differentiable at the
point Zo, and since Zo was arbitrary, f (z) is nowhere differentiable.
Remark
3.1
In Section 2.3 we showed that f (z) = z is continuous for all z.
Thus, we have a simple example of a function that is continuous everywhere but
differentiable nowhere. Such functions are hard to construct in real variables. In
some sense, the complex case has made pathological constructions simpler!
_
We seldom are interested in studying functions that aren't differentiable, or
are differentiable at only a single point. Complex functions that have a derivative
at all points in a neighborhood of Zo deserve further study. In Chapter 7 we
demonstrate that if the complex function f can be represented by a Taylor series
at zo, then it must be differentiable in some neighborhood of zoo Functions that
are differentiable in neighborhoods of points are pillars of the complex analysis
edifice; we give them a special name, as indicated in the following definition.
I Definition
3.1: Analytic
We say that the complex function f is analytic at the point zo, provided there
is some E > a such that l' (z) exists for all z E Dc: (zo). In other words, f must
be differentiable not only at zo, but also at all points in some E neighborhood of
zoo
If f is analytic at each point in the region R, then we say that f is analytic
on R. Again, we have a special term if f is analytic on the whole complex plane.
Definition
If
3.2: Entire
is said to be entire.
~zn
dz
:z
dd [j (z)
z
[C
0, where C
nznl,
1(z)] = C l'
+ g (z) 1= l'
dd [j (z)
z
d
dz
(z)] =
1(z)
g
(z)
a constant
and
(34)
(z) ,
I
(z) ,
(3 7)
+ g (z) l'
gl
(z)
l'
(z)  1(z)
[g (z)f
(z)
(35)
(:36)
+g
(z)
1(z)
.
IS
gI
(z),
(z)
(38)
provided that g (z) =F 0,
and
(39)
1(g
:z
dz zn
d
dz [j (z)t
(z))
l'
(g
(z))
gI
(310)
(z) .
zn+l'
= n [1(z)t1
EXAMPLE
l'
(z),
n a positive integer.
(;~12) with
1(z)
+ i2z + 3)4 =
8 (z2
The proofs of the rules given in Equations (34) through (310) depend on the
validity of extending theorems for real functions to their complex companions.
Equation (38), for example, relies on Theorem 3.1.
1is
continuous
at zo,
1(z)
 1(zo)
Z  Zo
z~zo
at zo, then
lim [j (z) 
1(zo)]
I' (zo)
z~zo
Z~ZO
,
11m
1(z)
z~zo
Z 
1(zo)
l'
(
)
1m z  zo
Zo
z+zo
=j'(zo)'O=O,
This result implies that
that
1 is continuous
lim
z+zo
1(z)
1(zo),
at zo,
Z4Z
h' (zo)
Z  Zo
Z4Z
(z)  1(zo)
Z  Zo
= 1(z)
(z)
(zo) ,
Z 
Z4Z
Zo
= lim 1(z) g (z)  1(zo) g (z) + lim 1(zo) g (z)  1(zo) g (zo)
Z  Zo
Z4Z
,
11m
z+zo
Z  Zo
Z4Z
Zo
2+Z0
Z+Zo
Z 
(zo) '
Zo
Using the definition of the derivative given by Equation (31) and the continuity
of g, we obtain h' (zo) = l' (zo) g (zo) + 1(zo) g' (zo), which is what we wanted
to establish. \Ne leave the proofs of the other rules as exercises.
The rule for differentiating polynomials carries over to the complex case as
well. If we let P (z) be a polynomial of degree n, so
~i~l
is analytic
f (z)
=I O.
z =I O. The
Arg(z)
that
2
1
2
2
f (z)
g(z)
f' (z) .
= lim
g' (z)
Z>ZQ
Proof
We defer the proof until Chapter
amazing things about analytic functions.
4z2
+ 7z
 8.
(b) g(z)=(z2iz+9)5.
2::21
(c) h(z)
(d) F (z)
= (Z2 +
(a) f (z)
(b)
for z
i 2.
(1  3i) z + 1)
Re (z) .
(z) = 1m (z) .
(a) [j (z)]3 .
(b) f(z)g(z).
(c)
at zoo
f(z).
g(z)
(d)f(~).
(e)f(zl).
(f) f (g (z)).
4. Use Equation
5. Let P (z) = ao
be a polynomial
of degree n
2:
1.
p(k) (0)
(k)
( b ) S how that, f01. k  0, 1, ... , n, ak ~ k!
,
where P
denotes the
kth derivative of P. (By convention, p(O) (z) = P (z) for all z.)
PI(z)
=+P(z)
ZZl
ZZ2'
Note: The quotient
7. Use L'Hopital's
(a) lim
as the logarithmic
derivative
of P.
z4_1.
zi
Z'L
()
I1m.
C
Z+'l.
lim
z+l+i
(e)
~'(Sz./ is known
z6+1
2+1'
~.
z 2z+2
z6 64
::J+8'
1m
z~l+iv'3
8. Use Equation
9. Show that
1.zn
= _nzn1,
1. ~ = ~.
~ f
f (z)
Izl2
+f
+f
12. Verify
(a) Identity
(:31).
(b) Identity
(37).
(c) Identity
(:39).
(d) Identity
010).
integer. Use the chain rule to show that if 9 (z) is any branch of the nth root
function, then
g' (z)
.!.g(z)
n z
in some suitably chosen domain (which you should specify).
15. Explain why the composition of two entire functions is an entire function.
16. Let
(z)
=f
(zo)
+ f ' (zo)
+ T) (z)
(z  zo)
T)
(z  zo) ,
1A.L. Cauchy (17891857) played a prominent role in the development of complex analysis,
and you will see his name several times throughout this text. The last name is not pronounced
as "kaushee." The beginning syllable has a long "0" sound, like the word kosher, but with the
second syllable having a long "e" instead of "er" at the end. Thus, we pronounce Cauchy as
"k6she."
1.0 1
4.06
Zo
t.
:       _. 
4.02
r
/,
,
O
We know that
3.02
W',"
3.04
is differentiable,
f (2 + i) ~ f (2.01 + i)  f (2 + i)
I
(2.01 + i)  (2 + i)
l' (2 + i) ~ f (2 + 1.01i)  f (2 + i)
(2 + 1.01i)  (2 + i)
0.0401 + 0.02i
0.01
4.01 + 2i
These computations lead to the idea of taking limits along the horizontal
and vertical directions. When we do so, we get
l' f (2 +
1m
f '(2 + Z.) = h>O
(2 + i) = lim
h>O
h + i) h
f (2 + i +
ih~ Z
f (2 + i)
h>O
f (2 + i)
4h + h + i2h
= 4 + 2i
h
= im 
= lim  2h  h
h>O
+ i4h
=4
+ 2i.
ih
differentiable
is differentiable
at the point Zo = Xo + iyo. Then the partial
and v exist at the point Xo + iyo = (xo, Yo), and
f' (zo)
= Ux
f' (zo)
= vy
Proof
Because
derivatives
of u
(314)
and also.
(315)
lim
Z*Zo
(f(Z)= f(zo)
z
exists
Zo
regardless of the path we take as z + zoo We will choose horizontal and
vertical lines that pass through the point Zo = (xo, Yo) and compute the
limiting values of f(;=~~)o) along these lines. Equating the two resulting
limits will yield Equations (316). For the horizontal approach to zo, we set
z = x + iyo and obtain
f' ()Zo=
1m
x  Xo
x>xo
lim _u_(x_,_y_o_)
__u_(_x_o_,
y_o_)+ i lim _v_(x_,_y_o_)
__v_(x_o_'_YO_)
.
X  Xo
x>xo
X  Xo
x>xo
The last two limits are the partial derivatives of u and v with respect to x,
so
f'()
Zo
Xo + iy, so
f(xo+iy)f(xo+iyo)
Xo + iy  (xo + iyo)
lim u(xO,y)u(xo,yo)+i[v(xo,Y)v(xO'yo)]
y>Yo
i (y  Yo)
l'
1m
(XO,Y)>(XO,YO)
lim _v_(x_o_,_y_)
__v_(_x_o_,
y_o_)_ i lim _u_(x_o_,_y_)
__u_(x_o_,y_o_).
y>Yo
y 
Yo
y>Yo
Y  Yo
The last two limits are the partial derivatives of u and v with respect to y,
so
l' (zo).
Taking the contrapositive, if Equations (316) are not satisfied at zo, then
we know automatically that f is not differentiable at zo.
Even if Equations (316) are satisfied at zo, we cannot necessarily
that f is differentiable at zoo
EXAMPLE 3.4
2z. We also have
We know that
f (z)
conclude
(x, y) = 2x = vy (x, y)
and
uy (x, y) = 2y =
Vx
(x, y) .
l' (z)
gives
EXAMPLE
3.5
Show that
f (z) = z
is nowhere differentiable.
Solution
(Z)2
f()
 z
=
{
X3 3x
x 2+ y
+2
. "+2
y3 3x2y
x y
whenz=lO
when z = 0
and
(0,0)
x  0
x3o
+o = l.
x2
x~o
lim
(x,o)~(O,o)
f (x + Oi)  f
x + Oi  0
(0)
= lim
x~o
x; X 
0
0
(t,t)~(O,O)
at Zo = O. Letting z approach 0
= lim x  0 = l.
x~o
=x
_2t + . (2t
f(t+it)
f(O)
2t2 22t2
= lim t + it  0
t~O
t + it
3
lim
equa
X 
equations
3
)
tit
=
t + it
= lim t~O
l.
Proof
Let 6.z = ~x + i6.y
enough so that z lies in the c
hold. We need to show that
(315) as 6.z approaches zero.
6.u
[u (xo
+ [u (xo,
+ 6.x,
The partial derivatives Ux and uy exist, so the mean value theorem for
real functions of two variables implies that a value x* exists between Xo and
Xo + 6.x such that we can write the first term in brackets on the right side
of Equation (317) as
Furthermore,
such that
as
Ux
C1
where
101 +
a as
6.x
+
and 6.y
+
O. Because 6.x
+
+
forces x*
+
such that the second term in brackets on the right side of Equation
satisfies the equation
where 102 +
(319) gives
as 6.x
+
and 6.y
+
O. Combining Equations
xo, we
102
(317)
(318) and
where partial derivatives Ux and uy are evaluated at the point (xo, Yo) and Cl
and C2 tend to zero as ~x and ~y both tend to zero. Similarly, the change
~v is related to the changes ~x and ~y by the equation
where th~ partial derivatives Vx and vy are evaluated at the point (xo, Yo)
and C3 and C4 tend to zero as ~x and ~y both tend to zero. Combining
these last two equations gives
~w
~w
hm ~=ux+wx+
.6.z+o
D.Z
[Cl~X
c2~y
hm
~+~+~~+~~
.6.z+o
D.Z
D.Z
.C3~X
D.Z
'C4~Y]
D.Z
.6.z+o D.Z
l' (zo)
= Ux
(xo, Yo)
+ ivx
C2,
(xo, Yo) ,
Solution
Vx
l' (z)
By Theorem
3 (x2  y2
(x, y)
Vx
(x, y) = uy
vY (x, y)
eY
sinx
= eYcosx
3.4,
+ i2xy)
+ i (3x2y
=

f is
3z2.
y3), and
and v(x,y)
and
We note that u, v, Ux, Uv, Vx, and vY are continuous functions and that the
CauchyRiemann equations hold for all values of (x, y). Hence, using Equation
(314), we write
sin x
+ ieY
cosx.
+ i (y3 + 3x2y)
is
Solution
Recall (Definition 3.1) that when we say a function is analytic at
a point Zo we mean that the function is differentiable not only at zo, but also
at every point in some E neighborhood of zoo With this in mind, we proceed
to determine where the Cauchy~Riemann equations are satisfied. We write
U (x, y) = x3 + 3xy2 and v (x, y) = y3 +3x2y
and compute the partial derivatives:
Ux
(x, y) = 3x2
uY (x, y)
6xy,
+ 3y2,
Vx
vY (x, y) = 3x2
(x, y)
6xy.
+ 3y2,
and
Here ux, uy, vx, and vy are continuous, and Ux (x, y) = vy (x, y) holds for all
(x, y). But uy (x, y) = Vx (x, y) iff 6xy = 6xy, which is equivalent to 12xy = O.
The CauchyRiemann equations hold only when x = 0 or y = 0, and according
to Theorem 3.4, f is differentiable only at points that lie on the coordinate axes.
But this means that f is nowhere analytic because any E neighborhood about a
point on either axis contains points that are not on those axes.
When polar coordinates (r, e) are used to locate points in the plane, we use
Expression (22) for a complex function for convenience; that is,
= u (x, y) + iv (x, y),
iB
f (re ) = u (reiB) + iv (TeiB) = U(T cos e, T sin e) + iV(T cos e, T sin e)
f (z)
= U(T, e)
+ iV
(T, e) ,
where U and V are real functions of the real variables T and e. The polar form
of the CauchyRiemann equations and a formula for finding l' (z) in terms of
the partial derivatives of U (T, e) and V (T, e) are given in Theorem 3.5, which
we ask you to prove in Exercise 10. This theorem makes use of the validity of
the CauchyRiemann equations for u and v, so the relation between them and
the functions U and Vnamely,
u(x,y)
= U(T,e) and v(x,y)
= V (T,e)is
important .
+ iV (T, e) be a
continuous function that is defined in some neighboThood of the point Zo =
ToeiBo. If all the paTtial deTivatives Ur, UB, Vr, and VB aTe continuous at the
point (ro, eo) and if the polaT form of the CauchyRiemann
equations,
1
1
TO
rO
Ur (ro, eo)
(ro, eo) ,
holds, then f is diffeTentiable at Zo and we can compute the derivative f' (zo)
by using eitheT
f' (zo)
f' (TeiBo)
f' (zo)
f' (TeiBo)
~eieo
+ iVr
(ro, eo)]
OT
(323)
(32<1)
TO
EXAMPLE
by
B)
f ( Te'
3.10
Show that if
lIe
=f(z)=z2=T2cos"2+iT2sin"2'
U(r,e)=r2cos"2
to be {re'ie : l'
> 0 and 
1f
Ur(r,e) = Ve(r,e)
l,e
= r2cos2
l'
1
l,e
and
l'
Since U, V, Un Ue, Vr, and Ve are continuous at every point in the domain (note
the strict inequality in 1f < < 1f), we use Theorem 3.5 and Equation (323)
to get
Note that f (z) is discontinuous on the negative real axis and is undefined at the
origin. Using the terminology of Section 2.4, the negative real axis is a branch
cut, and the origin is a branch point for this function.
Two important
section.
0, then it must be that u (x, y)2 = 0 and v (x, y)2 = 0 for all
f is identically zero on D. If K =I 0, then we take the partial
derivative of both sides of Equation (325) with respect to both x and y,
resulting in
If K
(x, y) ED, so
where for brevity we write u in place of u (x, y), and so on. We can now use
the CauchyRiemann equations to rewrite this system as
Ux
Ux
= u2 +v 2 =
and
uy
= u2 +v 2 = o.
in the domain D. If
l'
(z)
Proof
By the CauchyRiemann
equations, f' (z) = Ux (z) + ivx (z) =
vy (z)  iuy (z) for all zED.
By hypothesis f' (z) = 0 for all z in D, so for
all zED
the functions ux, uy, vx, and vy are identically zero. As with the
conclusion to the proof of Theorem 3.6, this situation means both u and v
are constant functions, from whence the result follows.
=iz+4i.
:,++iXy 2.
(b) f(z)=f(x,y)=
(c) f(z)=2(xy+x)+i(x22yy2).
(d) f (z) = x3  3x2
3xy2
+ 3y2 + i (3x2y
 6xy _ y3) .
(e) I(z)=x3+i(1y)3.
(f) I(z)~z2+z.
(g) I (z) = x2
(h)
I (z)
= Iz 
+ y2 + i2xy.
(2
+ i)12.
2. Let I be a differentiable
function.
+ iex
>
0,
sin y. Show that both I (z) and I' (z) are differentiable
(a)
I (z)
I is
I (z) =
= (lnr)2  82 + i2elnr,
analytic for r > 0, 'if
where r
>
>
0 and
find
'if
<
e:s:
'if.
+ i sinh x sin y.
I (z)
I (x,y)
I (reie)
U(r,e) +iV(r,8).
Use the transformation
(a) Let
(reie)
x = rcos8
U
iv (reiB)
and y = rsin8
Ur =
Ur
ax
Ux
Show
I' (z).
ar
+ Uy
= Ux cas 8
Vr = Vx cas e
ay
ar
+ uy sin e
+ vy sin 8
Ue
ax
Ux
ay
a8 + Uy a8
(similarly for V)
and
Ue
and
and
(b) Use the original CauchyRiemann equations for u and v and the results
of part (a) to prove that rUr = Ve and rVr = Ue, thus verifying
Equation (3~22)
(c) Use part (a) and Equations (314) and (3~15) to show that the right
sides of Equations (3~23) and (3~24) simplify to f (zo).
I
are differentiable
(z) = x3
(a)
(b)
f (z)
(c) f (z)
x2  y2
8y) .
+ i2lxyl.
12. Let f and 9 be analytic functions in the domain D. If l' (z) = 9 I (z) for all z in
D, then show that f (z) = 9 (z) + C, where C is a complex constant.
13. Explain how the limit definition for the derivative in complex analysis and the
limit definition for the derivative in calculus are different. How are they similar?
14. Let f be an analytic function in the domain D. Show that if Re
points in D, then f is constant in D.
[f (z)]
= 0 at all
15. Let f be a nonconstant analytic function in the domain D. Show that the function
9 (z) = f (z) is not analytic in D.
16. Recall that, for z
=x
+ iy,
x =
Z!Z
and y
= z;.z.
ah = ah ax
az
ax az
+ ah
ay = ~ (ah
ay az
2 ax
:z
+ i ah)
ay
= ~
(:x
+ i :y)
that
is suggested
by
~i
that
function
equations:
a domain
is a connected
open
~~
O.
x and y defined
set.)
The
partial
is known as Laplace's equation (sometimes referred to as the potential equation). If rP, rPx, rPy, rPxx, rPxy, rPyx, and rPyy are all continuous, and if rP(x, y) satisfies Laplace's equation, then rP (x, y) is harmonic on D. Harmonic functions are
important in applied mathematics, engineering, and mathematical physics. They
are used to solve problems involving steady state temperatures, twodimensional
electrostatics, and ideal fluid flow. In Chapter 11 we describe how complex analysis techniques can be used to solve some problems involving harmonic functions.
We begin with an important theorem relating analytic and harmonic functions.
Proof In Corollary 6.3 we will show that if f (z) is analytic, then all partial
derivatives of u and v are continuous. Using that result here, we see that,
as f is analytic, u and v satisfy the CauchyRiemann equations
Taking the partial derivative with respect to x of each side of these equations
gives
The partial derivatives uxy, uyx, vxy, and vyx are all continuous, so we use a
theorem from the calculus of real functions that states that the mixed partial
derivatives are equal; that is,
Combining all these results finally gives Uxx + Uyy = vyx  vxy = 0, and
+vyy = uyx +uxy = O. Therefore, both u and v are harmonic functions
Vxx
onD.
EXAMPLE 3.11
Ifu(x,y)
= X2_y2,
then Uxx (x,y)+uyy
(x,y) = 22 = 0;
hence u is a harmonic function for all z. We find that v (x, y) = 2xy is also a
harmonic function and that
of harmonic
EXAMPLE 3.12
The function f (z) = z3
analytic for all values of z; hence it follows that
x3  3xy2
+ i (3x2y
 y3) is
Figures 3.2 and 3.3 show the graphs of these two functions. The partial
derivatives are ux(x,y)
= 3x2  3y2, uy(x,y)
= 6xy,
vx(x,y)
= 6xy, and
vy (x, y) = 3x2  3y2. They satisfy the CauchyRiemann equations because they
are the real and imaginary parts of an analytic function. At the point (x, y) =
(2, 1), we have Ux (2, 1) = vy (2, 1) = 9, and these partial derivatives appear
along the edges of the surfaces for u and v where x = 2 and y = 1. Similarly,
uy (2, 1) = 12 and Vx (2, 1) = 12 also appear along the edges of the surfaces
for u and v where x = 2 and y = 1.
We can use complex analysis to show easily that certain combinations of
harmonic functions are harmonic. For example, if v is a harmonic conjugate of
u, then their product (x, y) = u (x, y) v (x, y) is a harmonic function. This condition can be verified directly by computing the partial derivatives and showing
that Equation .'](, holds, but the details are tedious. If we use complex variable techniques instead. we can start with the fact that f (z) = u (x, y) +iv (x, y)
is an analytic function. Then we observe that the square of f is also an analytic
function, which is
We then know immediately that the imaginary part, 2u (x, y) v (x, y), is a
harmonic function by Theorem 3.8. A constant multiple of a harmonic function
is harmonic, so it follows that is harmonic. We leave as an exercise to show
that if Ul and 'U2 are two harmonic functions that are not related in the preceding
fashion, then their product need not be harmonic .
jvy(x,Y)dY=
jux(x,Y)dY+C(X),
d~j
ux(x,y)dy+C'(x).
It can be shown (we omit the details) that because u is harmonic, all terms
except those involving x in the last equation will cancel, revealing a formula
To verify that u is harmonic, we compute the second partial derivatives and note
that Uxx (x, y) + uyy (x, y) = 6xy + 6xy = 0, so u satisfies Laplace's Equation
(326). To construct v, we start with Equation (327) and the first of Equations
(328) to get
Differentiating the left and right sides of this equation with respect to x and
using uy (x, y) = Vx (x, y) and Equations (328) on the left side yield
Harmonic functions arise as solutions to many physical problems. Applications include twodimensional models of heat flow, electrostatics, and fluid flow.
We now give an example of the latter.
We assume that an incompressible and frictionless fluid flows over the complex plane and that all cross sections in planes parallel to the complex plane are
Figure 3.4
fluid flow.
+ iq
the same. Situations such as this occur when fluid is flowing in a deep channel.
The velocity vector at the point (x. y) is
(x, y)
uy (x, y)
and
Assuming that the functions P and q have continuous partials, Theorem 3.4
guarantees that function f defined in Equation (331) is analytic and that the
fluid flow of Equation (329) is the conjugate of an analytic function; that is,
Theorem 3.8 implies that rP (x, y) is a harmonic function. Using the vector
interpretation of a complex number, the gradient of rP can be written as
Z2,
Figme
{(x,y)
The
3.;)
: 'l/J (x,y)
"Figure 3.6
families
= constant}
of orthogonal
curves
The equipotential
= eX cosy
(b) u (x, y)
In (x2
and v(x,y)
+ y2)
cP(x,y)
eXsiny.
= constant}
and
+i'l/J (x,y).
(a) u(x,y)
{(x,y):
= cP(x,y)
f (z)
+ iv (x, y)
u (x, y)
3. Let a, b, and c be real constants. Determine a relation among the coefficients that
will guarantee that the function (x, y) = ax2 + bxy + cy2 is harmonic.
4. Let v (x, y) = arctan ~ for x =1= O. Compute the partial derivatives of v, and verify
that v satisfies Laplace's equation.
5. Find an analytic function
(z) = u (x, y)
+ iv
= sinysinhx.
(c) v(x.y)
= eYsinx.
(d) v(x,y)
= sinxcoshy.
= x3 Ul
3xy2
(x, y)
U2
7. Assume that U (x, y) is harmonic on a region D that is symmetric about the line
y = O. Show that U(x, y) = U (x, y) is harmonic on D. Hint: Cse the chain rule
for differentiation of real functions and note that U (x, y) is really the function
u(g(x,y)),
where g(x,y)
= (x,y).
8. Let v be a harmonic conjugate of u. Show that u is a harmonic conjugate
9. Let v be a harmonic conjugate of u. Show that h
of v.
v is a harmonic function.
= (r+~)
(b) u (r, e)
= rn
= (r 
= rn
~)sine.
sin nB.
(x,y)
+i'lj;(x,y).
'lj; =
1, ~, ~.
Suppose that (xo, YO) is a point common to the two curves (x, y) = C1 and
'l/J (x, y) = C2 Use the gradients of and 'l/J to show that the normals to the curves
are perpendicular.
15. We introduce the logarithmic function in Chapter 5. For now, let F(z) = Logz =
Inlzl +iArgz.
Here we have (x,y)
= Inlzl and 'l/J(x,y)
= Argz.
Sketch the
equipotentials = 0, In 2, In 3, In 4 and the streamlines 'l/J = kS for k = 0, 1, ... , 7.
7C